Extended Data Fig. 4: A two-step recruitment approach for clinical trials in preclinical Alzheimer’s disease using the mPACC as outcome measure with plasma p-tau217 and Tau-PET_NEO.

Note that all analyses are performed using Tau-PET_NEO only. a, the obtained sample size reduction using sample selection based on different percentiles (75th, 50th and 25th) of baseline plasma p-tau217 levels in step 1 followed by the selection based on the same percentiles (75th, 50th and 25th) of the Tau-PET_NEO measurement in step 2 with mPACC5 as the primary endpoint. Note that 100% in step 2 refers to the participants selected by plasma p-tau217 in step 1. Error bars represent the 95% CI around the mean derived from mixed effects models. b, shows the calculated sample size reductions for various plasma p-tau217 and Tau-PET_NEO quartile combinations. Calculations are based on the assumption of 80% power to detect a 30% change on the mPACC5 in a 4-year trial. Red lines represent step 1 with plasma p-tau217 and blue lines represent step 2 with Tau-PET_NEO. Different line styles represent different quartiles of Tau-PET_NEO from those subjects already selected from step 1. Dotted black lines represent 100% participants needed without that step.